Merge branch 'romain-goodies' into unstable

gargantext/util/db.py

@@ -86,7 +86,19 @@ class bulk_insert:
     readline = read
 
 
-def bulk_insert_ifnotexists(model, uniquekey, fields, data, cursor=None):
+def bulk_insert_ifnotexists(model, uniquekey, fields, data, cursor=None, do_stats=False):
+    """
+    Inserts bulk data with an intermediate check on a uniquekey
+    (ex: Ngram.terms) to see if the row existed before.
+
+    If the row already existed we just retrieve its id.
+    If it didn't exist we create it and retrieve the id.
+
+    Returns a dict {uniquekey => id}
+
+    Option:
+        do stats: also returns the number of those that had no previous id
+    """
     if cursor is None:
         db, cursor = get_cursor()
         mustcommit = True
@@ -109,6 +121,7 @@ def bulk_insert_ifnotexists(model, uniquekey, fields, data, cursor=None):
         sourcetable = model.__tablename__,
         uniquecolumn = uniquekey,
     ))
+
     # insert what has not been found to the real table
     cursor.execute('''
         INSERT INTO {sourcetable} ({columns})
@@ -119,6 +132,11 @@ def bulk_insert_ifnotexists(model, uniquekey, fields, data, cursor=None):
         sourcetable = model.__tablename__,
         columns = ', '.join(fields),
     ))
+
+    if do_stats:
+        # remember how many rows we inserted just now
+        n_new = cursor.rowcount
+
     # retrieve dict associating unique key to id
     cursor.execute('''
         SELECT source.id, source.{uniquecolumn}
@@ -130,10 +148,15 @@ def bulk_insert_ifnotexists(model, uniquekey, fields, data, cursor=None):
         columns = ', '.join(fields),
     ))
     result = {
+        # term : new_id
         row[1]: row[0] for row in cursor.fetchall()
     }
     # this is the end!
     cursor.execute('DROP TABLE __tmp__')
     if mustcommit:
         db.commit()
-    return result
+
+    if do_stats:
+        return result, n_new
+    else:
+        return result

gargantext/util/group_tools.py

+"""
+Utilities for group management
+  - query_grouped_ngrams(group_id) to retrieve subforms
+  - group_union() to join two groupings lists
+"""
+from gargantext.util.db  import session, aliased
+from gargantext.models   import Ngram, NodeNgramNgram
+from igraph              import Graph  # for group_union
+
+def query_groups(groupings_id, details=False):
+    """
+    Listing of couples (mainform,   subform)
+                 aka   (ngram1_id, ngram2_id)
+
+    Parameter:
+      - details: if False, just send the array of couples
+                 if True, send quadruplets with (ngram1_id, term1, ngram2_id, term2)
+    """
+    if not details:
+        # simple contents
+        query = session.query(NodeNgramNgram.ngram1_id, NodeNgramNgram.ngram2_id)
+    else:
+        # detailed contents (id + terms)
+        Ngram1 = aliased(Ngram)
+        Ngram2 = aliased(Ngram)
+        query = (session
+                    .query(
+                        NodeNgramNgram.ngram1_id,
+                        Ngram1.terms,
+                        NodeNgramNgram.ngram2_id,
+                        Ngram2.terms,
+                     )
+                    .join(Ngram1, NodeNgramNgram.ngram1_id == Ngram1.id)
+                    .join(Ngram2, NodeNgramNgram.ngram2_id == Ngram2.id)
+                )
+
+    # main filter
+    # -----------
+    query = query.filter(NodeNgramNgram.node_id == groupings_id)
+
+    return query
+
+def query_grouped_ngrams(groupings_id, details=False, scoring_metric_id=None):
+    """
+    Listing of "hidden" ngram_ids from the groups
+
+    Works only for grouplists
+
+    Parameter:
+      - details: if False, send just the array of ngram_ids
+                 if True, send triples with (ngram_id, term, scoring)
+                                                             ^^^^^^^
+
+      deprecated: scoring_metric_id: id of a scoring metric node   (TFIDF or OCCS)
+                           (for details and sorting)
+                   (no more OCCS counts of subforms)
+    """
+    if not details:
+        # simple contents
+        query = session.query(NodeNgramNgram.ngram2_id)
+    else:
+        # detailed contents (terms and some NodeNodeNgram for score)
+        query = (session
+                    .query(
+                        NodeNgramNgram.ngram2_id,
+                        Ngram.terms,
+                        # NodeNodeNgram.score           #
+                     )
+                    .join(Ngram, NodeNgramNgram.ngram2_id == Ngram.id)
+                    # .join(NodeNodeNgram, NodeNgramNgram.ngram2_id == NodeNodeNgram.ngram_id)
+                    # .filter(NodeNodeNgram.node1_id == scoring_metric_id)
+                    # .order_by(desc(NodeNodeNgram.score))
+                )
+
+    # main filter
+    # -----------
+    query = query.filter(NodeNgramNgram.node_id == groupings_id)
+
+    return query
+
+
+def group_union(g_a_links, g_b_links):
+    """
+    Synonym groups are modelled by sets of couples in the DB
+
+    Input : 2 arrays of links (ngramx_id, ngramy_id)
+    Input : 1 array of links (ngramx_id, ngramy_id)
+
+    Synonymity is considered transitive so in effect the groups
+    can form a set (defined by the connected component of couples).
+
+     A requested feature is also that one node dominates others
+     (aka "leader effect"; leader will be in the map, the others won't)
+
+    Summary of major union effects in various cases:
+
+    GROUP 1         Group 2         Group 1 ∪ 2
+
+    A -> B           A -> C           A -> B       (simple union)
+                                      A -> C
+
+    D -> E           E -> F           D -> E
+                                      D -> F       (D "leader effect")
+
+
+    G -> H           G -> I           G -> H       ( transitivity +
+                     H -> J           G -> I        "leader effect")
+                                      G -> J
+
+     rloth: this is some slightly amended code
+     from Samuel's in rest_v1_0.ngrams.Group.get
+
+     TODO use "most frequent" score if leader candidates are ex aequo by degree.
+    """
+
+    # output: list of links forming new group
+    new_links = []
+
+    # 1) create graph with both lists
+    # -------------------------------
+
+    # from igraph import Graph
+
+    # the set of all our ngram_ids
+    all_vertices = set(
+      [ngid for couple in g_a_links+g_b_links for ngid in couple]
+    )
+
+    # initialize the synonym graph with size
+    sg = Graph(len(all_vertices), directed=True)
+
+    # add our IDs as "name" (special attribute good for edge creation)
+    sg.vs['name'] = [str(x) for x in all_vertices]
+
+    # add the edges as named couples
+    sg.add_edges([(str(x),str(y)) for (x,y) in g_a_links])
+
+    #print('UNION A:', g_a_links)
+    #print('initially %i components' % len(sg.as_undirected().components()))
+
+    # same with the other edges
+    sg.add_edges([(str(x),str(y)) for (x,y) in g_b_links])
+
+    #print('UNION B:', g_b_links)
+    #print('after union %i components' % len(sg.as_undirected().components()))
+
+
+    # 2) list resulting components
+    # -----------------------------
+    synonym_components = sg.as_undirected().components()
+
+    # for example
+    # cs = [[0, 3, 6], [1, 2, 8], [4, 5, 9, 11], [7,10]]
+
+    # there should be no singletons by construction
+
+    # list of all outdegrees for "leader" detection
+    # (leader = term most often marked as source by the users)
+    odegs = sg.outdegree()
+
+    #for i, v in enumerate(sg.vs):
+    #    print("%i - name:%s - odeg:%i" % (i, v['name'], odegs[i]))
+
+    for component in synonym_components:
+        # we map back to our ids, preserving order
+        our_comp = [int(our_id) for our_id in sg.vs[component]['name']]
+
+        # 3) take main node and unnest into new links list
+        # -------------------------------------------------
+
+        # position (within this component) of the best node (by degree)
+        max_odeg = -1
+        main_node_local_index = None
+        for position, vertex_id in enumerate(component):
+            this_odeg = odegs[vertex_id]
+            if this_odeg > max_odeg:
+                main_node_local_index = position
+                max_odeg = this_odeg
+
+        # we set it aside in our translated version our_comp
+        main_node = our_comp.pop(main_node_local_index)
+
+        # and unnest the others
+        for remaining_id in our_comp:
+            new_links.append((main_node, remaining_id))
+
+    return new_links

gargantext/util/toolchain/ngrams_extraction.py

@@ -9,6 +9,9 @@ from re          import sub
 from gargantext.util.scheduling import scheduled
 
 def _integrate_associations(nodes_ngrams_count, ngrams_data, db, cursor):
+    """
+    @param ngrams_data   a set like {('single word', 2), ('apple', 1),...}
+    """
     print('INTEGRATE')
     # integrate ngrams
     ngrams_ids = bulk_insert_ifnotexists(

gargantext/views/api/ngramlists.py

@@ -8,118 +8,88 @@ API views for advanced operations on ngrams and ngramlists
 """
 
 from gargantext.util.http     import APIView, get_parameters, JsonHttpResponse,\
-                                     ValidationException, Http404
-from gargantext.util.db       import session, aliased, desc, bulk_insert
+                                     ValidationException, Http404, HttpResponse
+from gargantext.util.db       import session, aliased, bulk_insert
 from gargantext.util.db_cache import cache
 from sqlalchemy               import tuple_
 from gargantext.models        import Ngram, NodeNgram, NodeNodeNgram, NodeNgramNgram
 from gargantext.util.lists    import UnweightedList, Translations
 
+# useful subroutines
+from gargantext.util.ngramlists_tools import query_list, export_ngramlists, \
+                                             import_ngramlists, merge_ngramlists
+from gargantext.util.group_tools      import query_grouped_ngrams
 
-def _query_list(list_id,
-                pagination_limit=None, pagination_offset=None,
-                details=False, scoring_metric_id=None
-                ):
+
+class List(APIView):
+    """
+    see already available API query api/nodes/<list_id>?fields[]=ngrams
     """
-    Paginated listing of ngram_ids in a NodeNgram lists.
+    pass
 
-    Works for a mainlist or stoplist or maplist (not grouplists!)
 
-    Parameter:
-      - pagination_limit, pagination_offset
-      - details: if False, send just the array of ngram_ids
-                 if True, send triples with (ngram_id, term, scoring)
-                                                             ^^^^^^^
-      - scoring_metric_id: id of a scoring metric node   (TFIDF or OCCS)
-                           (for details and sorting)
+class CSVLists(APIView):
     """
-    if not details:
-        # simple contents
-        query = session.query(NodeNgram.ngram_id).filter(NodeNgram.node_id == list_id)
-    else:
-        # detailed contents (terms and some NodeNodeNgram for score)
-
-        # NB: score can be undefined (eg ex-subform that now became free)
-        #     ==> we need outerjoin
-        #     and the filter needs to have scoring_metric_id so we do it before
-        ScoresTable = (session
-                        .query(NodeNodeNgram.score, NodeNodeNgram.ngram_id)
-                        .filter(NodeNodeNgram.node1_id == scoring_metric_id)
-                        .subquery()
-                        )
-
-        query = (session
-                    .query(
-                        NodeNgram.ngram_id,
-                        Ngram.terms,
-                        ScoresTable.c.score
-                     )
-                    .join(Ngram, NodeNgram.ngram_id == Ngram.id)
-
-                    # main filter ----------------------
-                    .filter(NodeNgram.node_id == list_id)
-
-                    # scores if possible
-                    .outerjoin(ScoresTable,
-                               ScoresTable.c.ngram_id == NodeNgram.ngram_id)
-
-                    .order_by(desc(ScoresTable.c.score))
-                )
+    For CSV exports of all lists of a corpus
 
-    if pagination_limit:
-        query = query.limit(pagination_limit)
+    Or CSV import into existing lists as "patch"
+    """
+    def get(self, request):
+        params = get_parameters(request)
+        corpus_id = int(params.pop("corpus"))
+        corpus_node = cache.Node[corpus_id]
 
-    if pagination_offset:
-        query = query.offset(pagination_offsets)
+        # response is file-like + headers
+        response = HttpResponse(content_type='text/csv')
+        response['Content-Disposition'] = 'attachment; filename="corpus-%i_gargantext_term_list.csv"' % corpus_id
 
-    return query
+        # fill the response with the data
+        export_ngramlists(corpus_node, fname=response, titles=True)
+        return response
 
+    def post(self,request):
+        """
+        Merge the lists of a corpus with other lists from a CSV source
+                                                 or from another corpus
 
+        params in request.GET:
+            corpus:    the corpus whose lists are getting patched
 
+        params in request.FILES:
+            csvsource: the csv file
 
-def _query_grouped_ngrams(groupings_id, details=False, scoring_metric_id=None):
-    """
-    Listing of "hidden" ngram_ids from the groups
+        or in get
+            dbsource:  another corpus instead of the csvfile
+                       (? this last option should perhaps not be in CSVLists ?)
 
-    Works only for grouplists
+        NB: not using PATCH because we'll need POST file upload
 
-    Parameter:
-      - details: if False, send just the array of ngram_ids
-                 if True, send triples with (ngram_id, term, scoring)
-                                                             ^^^^^^^
 
-      deprecated: scoring_metric_id: id of a scoring metric node   (TFIDF or OCCS)
-                           (for details and sorting)
-                   (no more OCCS counts of subforms)
-    """
-    if not details:
-        # simple contents
-        query = session.query(NodeNgramNgram.ngram2_id)
-    else:
-        # detailed contents (terms and some NodeNodeNgram for score)
-        query = (session
-                    .query(
-                        NodeNgramNgram.ngram2_id,
-                        Ngram.terms,
-                        # NodeNodeNgram.score           #
-                     )
-                    .join(Ngram, NodeNgramNgram.ngram2_id == Ngram.id)
-                    # .join(NodeNodeNgram, NodeNgramNgram.ngram2_id == NodeNodeNgram.ngram_id)
-                    # .filter(NodeNodeNgram.node1_id == scoring_metric_id)
-                    # .order_by(desc(NodeNodeNgram.score))
-                )
+        /!\ We assume we checked the file size client-side before upload
 
-    # main filter
-    # -----------
-    query = query.filter(NodeNgramNgram.node_id == groupings_id)
+        £TODO check authentication and user.id
+        """
+        # this time the corpus param is the one with the target lists to be patched
+        params = get_parameters(request)
+        corpus_id = int(params.pop("onto_corpus"))
+        corpus_node = cache.Node[corpus_id]
 
-    return query
+        # request also contains the file
+        # csv_file has type django.core.files.uploadedfile.InMemoryUploadedFile
+        #                                                 ----------------------
+        csv_file = request.data['csvfile']
+
+        # import the csv
+        new_lists = import_ngramlists(csv_file)
+        del csv_file
+
+        # merge the new_lists onto those of the target corpus
+        log_msg = merge_ngramlists(new_lists, onto_corpus=corpus_node)
+
+        return JsonHttpResponse({
+            'log': log_msg,
+            }, 200)
 
-class List(APIView):
-    """
-    see already available API query api/nodes/<list_id>?fields[]=ngrams
-    """
-    pass
 
 
 class GroupChange(APIView):
@@ -441,7 +411,7 @@ class MapListGlance(APIView):
         listmembers = {'maplist':[]}         # ngram ids sorted per list name
 
         # infos for all ngrams from maplist
-        map_ngrams = _query_list(maplist_id, details=True,
+        map_ngrams = query_list(maplist_id, details=True,
                                       scoring_metric_id= scores_id).all()
 
         # ex:  [(8805, 'mean age', 4.0),
@@ -566,25 +536,25 @@ class ListFamily(APIView):
         if "head" in parameters:
             # head <=> only mainlist AND only k top ngrams
             glance_limit = int(parameters['head'])
-            mainlist_query = _query_list(mainlist_id, details=True,
+            mainlist_query = query_list(mainlist_id, details=True,
                                           pagination_limit = glance_limit,
                                           scoring_metric_id= scores_id)
         else:
             # infos for all ngrams from mainlist
-            mainlist_query = _query_list(mainlist_id, details=True,
+            mainlist_query = query_list(mainlist_id, details=True,
                                           scoring_metric_id= scores_id)
             # infos for grouped ngrams, absent from mainlist
-            hidden_ngrams_query = _query_grouped_ngrams(groups_id, details=True,
+            hidden_ngrams_query = query_grouped_ngrams(groups_id, details=True,
                                           scoring_metric_id= scores_id)
 
             # infos for stoplist terms, absent from mainlist
-            stop_ngrams_query = _query_list(other_list_ids['stoplist'], details=True,
+            stop_ngrams_query = query_list(other_list_ids['stoplist'], details=True,
                                             scoring_metric_id=scores_id)
 
             # and for the other lists (stop and map)
             # no details needed here, just the member ids
             for li in other_list_ids:
-                li_elts = _query_list(other_list_ids[li], details=False
+                li_elts = query_list(other_list_ids[li], details=False
                                       ).all()
                 # simple array of ngram_ids
                 listmembers[li] = [ng[0] for ng in li_elts]

gargantext/views/api/urls.py

@@ -27,6 +27,15 @@ urlpatterns = [ url(r'^nodes$'                , nodes.NodeListResource.as_view()
                 #                     \
                 #                   corpus id
 
+              , url(r'^ngramlists/export$', ngramlists.CSVLists.as_view()            )
+                # get a CSV export of the ngramlists of a corpus
+                #  ex: GET ngramlists/export?corpus=43
+                #  TODO : unify to a /api/ngrams?formatted=csv
+                #        (similar to /api/nodes?formatted=csv)
+
+              , url(r'^ngramlists/import$', ngramlists.CSVLists.as_view()            )
+                # same handling class as export (CSVLists)
+                # but this route used only for POST + file
 
               , url(r'^ngramlists/change$', ngramlists.ListChange.as_view()          )
                 # add or remove ngram from a list

gargantext/views/pages/terms.py

@@ -33,6 +33,9 @@ def ngramtable(request, project_id, corpus_id):
             'project': project,
             'corpus' : corpus,
             'resourcename' : resourcename(corpus),
-            'view': 'terms'
+            'view': 'terms',
+
+            # for the CSV import modal
+            'csvimportroute': "/api/ngramlists/import?onto_corpus=%i"% corpus.id
         },
     )

install/python/requirements.txt

@@ -11,6 +11,7 @@ django-pgfields==1.4.4
 django-pgjsonb==0.0.16
 djangorestframework==3.3.2
 html5lib==0.9999999
+python-igraph>=0.7.1
 jdatetime==1.7.2
 kombu==3.0.33                  # messaging
 nltk==3.1

static/lib/gargantext/menu.css

@@ -19,3 +19,13 @@
     line-height: .85;
     margin-bottom: -5px;
 }
+
+.exportbtn {
+    /* border: 1px solid #333 ; */
+    margin-top:17px ;   /* valigns with bootstrap h2  */
+}
+
+.btn .glyphicon {
+    /* glyphicons are always rendered too high within bootstrap buttons */
+    vertical-align:middle
+}

templates/pages/corpora/terms.html

@@ -72,6 +72,15 @@
                           <button id="Save_All" class="btn btn-muted" disabled style="font-size:120%">
                               <b>Save all changes</b>
                           </button>
+                          <br/>
+                          <br/>
+                          <!-- import icon -->
+                          <span class="needsaveicon glyphicon glyphicon-import"></span>
+                          &nbsp;
+                          <button id="ImportList" class="btn btn-warning" style="font-size:120%"
+                                   onclick="$('#csvimport').modal('show');">
+                              <b>Import a Termlist</b>
+                          </button>
                       </div>
                         <!-- see in javascript function queries.functions['my_state_filter'] -->
                         <div class="pull-right" style="margin-top:2.1em;padding-left:1em;">
@@ -107,25 +116,110 @@
                 </div> <!-- /div panel -->
 
         </div> <!-- /jumbotron -->
-    <!--
-    <button id="ImportList" onclick="GetUserPortfolio();" class="btn btn-warning">
-        Import a Corpus-List
-    </button>
--->
-<!--</div> This div is closed in the menu !-->
-
 
+<!--</div> This div is closed in the menu !-->
 
 
-    <!--
-    # stub to import a list (aka orange button)
-    <button id="ImportList" onclick="GetUserPortfolio();" class="btn btn-warning">Import a Corpus-List</button>
-    -->
 
+<div class="modal" aria-hidden="true" id="csvimport">
+    <div class="modal-dialog">
+    <div class="modal-content">
+    <div class="modal-header">
+     <button type="button" class="close" data-dismiss="modal" aria-hidden="true">×</button>
+     <h3 id="myModalLabel">Import a CSV term list</h3>
+    </div>
+    <div class="modal-body" id="uploadform">
+        <form id="csvimportform"
+              onsubmit="return postCSV(event)"
+              enctype="multipart/form-data"
+              method="post">
+            {% csrf_token %}
+            <label>From your disk:</label>
+            <input type="file" id="csvfile" accept="text/csv">
+            <br/>
+            <label>From another corpus:</label>
+            <p>TODO</p>
+            <br/>
+            <input type="submit" class="btn btn-xs btn-info" id="csvsubmit" value="Submit" />
+        </form>
+    </div>
+    <div class="modal-footer" id="formanswer"></div>
+    </div>
+    </div>
+</div>
 
 <script type="text/javascript" src="{% static "lib/jquery/dynatable/jquery.dynatable.js" %}"></script>
 <!-- custom-lib for dynatable.js and dc.js -->
 <script type="text/javascript" src="{% static "lib/gargantext/NGrams_dyna_chart_and_table.js" %}"></script>
 
+<script type="text/javascript">
+
+/*                                             merci c24b !
+ * Uses csvimportroute variable from the django template
+ * Ex: /api/ngramlists/import?onto_corpus=corpus_id
+ *
+ * Uses input#csvfile as source data.
+ */
+function postCSV(e){
+    // don't do page reload of usual submits
+    e.preventDefault()
+
+    // 2MB ≈ 70000 ngrams
+    var max_size = 2097152
+
+    // we take it straight from the input element
+    theFile = $('input#csvfile')[0].files[0]
+
+    // debug
+    // console.log(theFile.name, "size", theFile.size, theFile.lastModifiedDate)
+
+    if (! theFile) {
+        console.warn('Ignoring "submit": no provided file')
+        return false
+    }
+    else if (theFile.size > max_size) {
+        console.warn('Ignoring "submit": file is too big')
+        $('#formanswer').html(
+            'The import failed: your file is too big ('+max_size/1024+'kB max).'
+        );
+        return false
+    }
+    // normal case
+    else {
+        // append into an empty form (or fixme: initialize it using form element)
+        var myFileFormData = new FormData();
+        myFileFormData.append("csvfile", theFile)
+
+        //postCorpusFile
+        $.ajax({
+             url: "{{csvimportroute | safe}}",
+             type: 'POST',
+             async: true,
+             contentType: false,
+             processData: false,
+             data: myFileFormData,
+             beforeSend: function(xhr) {
+               xhr.setRequestHeader("X-CSRFToken", getCookie("csrftoken"));
+             },
+             success: function(response) {
+                 my_html  = "<h2 color='green'>IMPORT OK ! </h2>"
+                 my_html += "<p class='note'>" + response['log'].replace(/\n/g, '<br/>') + "</p>"
+                 my_html += "<p'>(this page will reload in 3s)</p>"
+                 $('#formanswer').html(my_html);
+                 console.log(response) ;
+                 // reload after 3s
+                 setTimeout("location.reload(true)", 3000);
+                 },
+              error: function(result) {
+                  $('#formanswer').html('Erreur');
+                  console.error(result);
+                 },
+            });
+        $('#formanswer').html('CSV import in Progress');
+    }
+};
+</script>
+
+
 
 {% endblock %}

templates/pages/menu.html

@@ -41,7 +41,7 @@
                         {% if corpus %}
                         <li><a href="/projects/{{project.id}}/corpora/{{corpus.id}}">
                                 <span class="glyphicon glyphicon-file" aria-hidden="true"></span>
-                                {{corpus.name | truncatechars:15}}
+                                {{corpus.name | truncatechars:25}}
                             </a>
                         </li>
                         {% endif %}
@@ -150,12 +150,32 @@
                             <br>
                             <br>
                                 <div class="row">
-                                    <h3>
-                                        <a href="/projects/{{project.id}}">
-                                            <span class="glyphicon glyphicon-book" aria-hidden="true"></span>
-                                            {{ project.name | truncatechars:50}}
+                                    <div class="col-md-6">
+                                        <h3>
+                                            <a href="/projects/{{project.id}}">
+                                                <span class="glyphicon glyphicon-book" aria-hidden="true"></span>
+                                                {{ project.name | truncatechars:50}}
+                                            </a>
+                                        </h3>
+                                    </div>
+                                    <!-- export button -->
+                                    <div class="col-md-6">
+                                        {% if view == 'terms'  %}
+                                        <a class="btn btn-primary exportbtn pull-right" role="button"
+                                            href="/api/ngramlists/export?corpus={{corpus.id}}"
+                                            title="Export terms table in CSV">
+                                            Export terms table &nbsp; <span class="glyphicon glyphicon-download" aria-hidden="true"></span>
+                                        </a>
+                                        {% elif view == 'titles'  %}
+                                        <a class="btn btn-primary exportbtn pull-right" role="button"
+                                            href="/api/nodes?parent_id={{corpus.id}}&types[]=DOCUMENT&pagination_limit=100000&formated=csv"
+                                            title="Export full corpus in CSV">
+                                            Export corpus &nbsp; <span class="glyphicon glyphicon-download" aria-hidden="true"></span>
                                         </a>
-                                    </h3>
+                                        {% else  %}
+                                        <!-- TODO export journal table -->
+                                        {% endif %}
+                                    </div>
                                 </div>
                                 <div class="row">
                                     <div class="col-md-1">
@@ -167,10 +187,7 @@
                                             </h3>
                                             <h3>
                                                 <span class="glyphicon glyphicon-file" aria-hidden="true"></span>
-                                                {{ corpus.name | truncatechars:20 }}
-                                                <a class="btn btn-primary" role="button" href="/api/nodes?parent_id={{corpus.id}}&types[]=DOCUMENT&pagination_limit=100000&formated=csv">
-                                                    <span class="glyphicon glyphicon-download" aria-hidden="true"></span>
-                                                </a>
+                                                {{ corpus.name | truncatechars:30 }}
                                             </h3>
 
                                     </div>